A Novel Technique for Detecting Concealed Malicious URLs within the Tor Network

Onion URLs lead to the dark web, a mysterious and secretive internet space with many websites. This paper proposes a novel content-based classification of. onion URLs. Given the concerns surrounding the dark web’s anonymity, the project aims to classify these hidden services as either ‘benign’ or ‘potentially harmful. ’ The method involves scraping. onion websites for text and metadata, followed by natural language processing and machine learning analysis of word frequencies, sentiment, and topic modeling. A labeled dataset with strict criteria is utilized to train and fine-tune classification models. This project assists dark web users in identifying legitimate and risky services and contributes to discussions on content analysis and classification in anonymized online environments. With onion URLs mostly hidden and unindexed, the research advances privacy and security in this unique digital space. By automating the classification of. onion URLs, a valuable tool is provided, contributing to content analysis, online anonymity, and internet safety discussions. The problem statement seeks a Proof of Concept (PoC) to list active TOR-hosted hidden server URLs (.onion).

Hali tarjima qilinmagan